Head and neck squamous cell carcinoma (HNSCC) is the sixth most frequent cancer worldwide, with only a 50% cure rate in spite of combined treatment modalities. Therapeutic targeting of the epidermal growth factor receptor (EGFR) improves the survival in a subset of patients, although molecular predictors of sensitivity are currently elusive. Moreover, responsive patients often acquire resistance and ultimately succumb to their disease. Distinguishing the specific molecular processes that drive such therapeutic resistance amid complex cross-talk in cell signaling processes and stochastic evolutionary pressures requires dynamical models built from serial data. Therefore, in this application, we develop novel computational algorithms to infer the molecular mechanisms underlying cetuximab resistance from in vitro and in vivo model of cetuximab resistant HNSCC. Specifically, we will investigate the hypotheses that: (1) short-term time course data improve the ability of in silicon modeling techniques to infer both on- and off-target signaling responses to cetuximab; (2) combined epigenetic, post-transcriptional, and genomic changes in HNSCC cells upon chronic exposure to cetuximab result in acquired resistance; and (3) modeling inter and intra-individual heterogeneity will discern the specific cellular signaling processes that are activated to drive in vivo acquired cetuximab resistance in cell- line xenograft models of HNSCC. The results from this project will ultimately contribute to the selection of patients for cetuximab treatment and alternative molecular targets to overcome acquired cetuximab resistance. The algorithms developed will also be directly applicable to inference of molecular drivers of therapeutic resistance in additional cancers.